Supplement of Molecular Insights on Aging and Aqueous-Phase Processing from Ambient Biomass Burning Emissions-Influenced Po Valley Fog and Aerosol
Total Page:16
File Type:pdf, Size:1020Kb
Supplement of Atmos. Chem. Phys., 18, 13197–13214, 2018 https://doi.org/10.5194/acp-18-13197-2018-supplement © Author(s) 2018. This work is distributed under the Creative Commons Attribution 4.0 License. Supplement of Molecular insights on aging and aqueous-phase processing from ambient biomass burning emissions-influenced Po Valley fog and aerosol Matthew Brege et al. Correspondence to: Lynn R. Mazzoleni ([email protected]) The copyright of individual parts of the supplement might differ from the CC BY 4.0 License. Table of Contents 1. FT-ICR MS data processing and molecular formula assignment review 2. Ultrahigh resolution FT-ICR MS results 3. FT-ICR MS data set 4. Supplemental figures and tables a. Figure S1: Distribution of the molecular formulas by elemental group subclasses b. Figure S2: van Krevelen diagrams of all assigned molecular formulas scaled to number of oxygen atoms c. Figure S3: Reconstructed difference mass spectra of assigned molecular formulas d. Figure S4: Oxygen difference of abundance trends e. Figure S5: Double bond equivalent difference of abundance trends f. Figure S6: Kendrick mass defect diagrams of assigned molecular formulas scaled to number of oxygen atoms for unique molecular formulas per sample g. Figure S7: Kendrick mass defect diagrams of all assigned molecular formulas scaled to number of oxygen atoms h. Table S1: Summary of possible identified molecular formulas 1. FT-ICR MS data processing and molecular formula assignment review The individual transient scans of FT-ICR MS data for each sample were reviewed manually and the unacceptable scans with an abrupt change in the total ion current were removed; the remaining transient scans were co-added together to create the working file for each sample (this helped to increase signal to noise and enhance sensitivity). Molecular formula assignments were made as previously described (Mazzoleni et al., 2010; Putman et al., 2012; Zhao et al., 2013; Dzepina et al., 2015) using Sierra Analytics Composer software (version 1.0.5) within the limits of: C2-200H4-1000O1-20N0-3S0-1. Masses were calculated from measured m/z values, assuming an ion charge of -1 from the electrospray. The calculator uses a CH2 Kendrick mass defect (KMD) analysis to sort homologous ion series and extend the molecular formula assignments to higher masses (Hughey et al., 2001; Kujawinski and Behn, 2006). A de novo cut-off at m/z 500 was applied, indicating that no new formula assignments would occur above m/z 500, unless the formula was part of an existing CH2 homologous series that began at a point lower than m/z 500. This is necessary because the number of possible molecular formulas increases at higher values. The minimum relative abundance required for molecular formula assignment was > 10 times the estimated signal-to-noise ratio, determined for each sample between m/z 900–1000. Only integer values up to 40 were allowed for the double bond equivalents (DBE). The data set was manually reviewed to remove: formulas with an absolute error > 3 ppm, elemental ratios that were not chemically sensible (such as O:C > 3 or H:C < 0.3), and formulas which violated the rule of 13 or violated the nitrogen rule. The rule of 13 checks for a reasonable number of heteroatoms in a formula. A base formula (CnHn+r) 푀 푟 can be generated for any measured mass by solving: = 푛 + (Pavia, 2009). Then, the maximum number of "large 13 13 atoms" (C, O, N, S) in a formula is defined as the mass divided by 13, because substituting for a heteroatom (O, N or S) involves a substitution for at least one carbon. This maximum number is then compared to the actual number of "large atoms" in a formula, and those formulas exceeding the maximum number are rejected. The nitrogen rule removes formulas with odd masses that do not contain an odd number of nitrogen atoms, and even masses that do not contain an even number (or no) nitrogen atoms; this is due to the odd numbered valence of nitrogen (Pavia, 2009). Molecular formulas that contained 13C or 34S were also removed from the data set. Homologous series with large gaps in the DBE trend were removed, as well as homologous series with a length of one. The assigned formulas were also analyzed with consideration to the DBE and oxygen number trends, (Herzsprung et al., 2014) where unreliable formula assignments were also removed. 2. Ultrahigh resolution FT-ICR MS results The total ion abundance of the identified monoisotopic molecular formulas reported for each sample was determined by their summation. Then, these values were used to normalize the individual ion abundances within each sample using a ratio of the individual ion intensity to this total ion abundance. Then, the values were rescaled using a normalization constant (10,000). This normalization procedure was done to remove analytical biases introduced by trace contaminants with high electrospray efficiency. Reconstructed difference mass spectra of the assigned molecular formulas for both fog and aerosol samples are shown in Fig. S3. These difference mass spectra permit a direct comparison of the samples using normalized S2 relative abundances. The individual relative abundances were normalized by the total abundance of the assigned molecular formulas identified in each of the samples. In Fig. S3, the individual masses with higher abundances in either the positive or negative direction were substantially greater in one of the two samples, whereas the masses of similar abundance tended to cancel each other. To enhance the interpretation of the compositional differences, the individual masses were color-coded to represent the number of oxygen atoms in the assigned formula. Overall, we observed higher numbers of oxygen in the masses of the two samples with aged biomass burning emissions influence compared to the two samples with fresh biomass burning emissions influence. The molecular formulas assigned to the fresh samples had approximately 0-5 oxygen atoms over the mass range of 50-250 Da, 5-10 oxygen atoms over 250- 550 Da, and a few molecular formulas were assigned with 10-15 oxygen atoms over 500-600 Da. In contrast, the aged samples had a large number of molecular formulas with 10-15 oxygen atoms in the range of 400-550 Da. This clearly shows a greater amount of oxidation in the aged influenced samples compared to the fresh influenced samples. KMD diagrams can be used as useful tools to visualize the relationships between the many molecular formulas of complex mixtures such as atmospheric samples. We used Kendrick mass defect to sort the molecular formulas into CH2 homologous series of identical heteroatom content and DBE, where the formulas in the same series differ only by a number of CH2 units (Stenson et al., 2003). It should be noted that the presence of multiple formulas in the same homologous series does not necessarily imply a related chemical structure. The homologous series are visible as horizontal rows of formulas in Figs. S6 and S7. There were multiple homologous series per subclass, where the base formula for each series differ in DBE and increase in KMD to form an ensemble of “steps” within each subclass. In our samples individual CHO and CHNO subclasses had approximately 5-16 different homologous series, while CHOS and CHNOS subclasses had approximately 3-10 different homologous series. The number of homologous series in a subclass increased with oxygen number, and peaked near the median oxygen number, then decreased again towards the maximum number of oxygen; this led to fewer molecular formulas in subclasses with higher and lower oxygen numbers, and more formulas in subclasses near the median oxygen number. The subclasses with the highest numbers of molecular formulas per elemental group were: O7, NO8, O7S and NO9S. It was atypical for the unique formulas of a sample to be completely unrelated to other formulas across the data set; often the unique formulas were extensions of homologous series that appeared across samples. 3. FT-ICR MS data set An abbreviated list of the complete FT-ICR MS dataset is provided and is available on Digital Commons: http://digitalcommons.mtu.edu/chemistry-fp/98/ S3 4. Supplemental Figures and tables Figure S1: Distributions of the molecular formulas within all 64 elemental group subclasses for CHO, CHNO, CHOS and CHNOS groups as indicated in the Figure. The total number of molecular formulas for each SPE-recovered WSOC sample were split into two groups of unique and non-unique formulas; the darker shade represents formulas unique to a sample, (denoted in the Figure legend with an asterisk after the sample name, e.g. “Fresh Fog*”) while the lighter shade represents common formulas. The sample names Fresh Fog, Aged Fog, Fresh Aerosol, and Aged Aerosol correspond to SPC0106F, SPC0201F, BO0204N, and BO0213D, respectively. S4 Figure S2: van Krevelen diagrams for the SPE-recovered WSOC by elemental group (rows) and sample (columns) as indicated in the Figure. Dashed lines represent H:C = 1.2 (horizontal), O:C = 0.6 (vertical) and OSC = 0 (diagonal) as described in Tu et al. (2016). Formulas are color scaled to the number of oxygen atoms in the assigned formula. The sample names Fresh Fog, Aged Fog, Fresh Aerosol, and Aged Aerosol correspond to SPC0106F, SPC0201F, BO0204N, and BO0213D, respectively. S5 Figure S3: Reconstructed difference mass spectra for theoretical masses of assigned molecular formulas in the Po Valley samples with normalized relative abundance. Fresh influenced samples (SPC0106F and BO0204N) are plotted with positive abundance and aged influenced samples (SPC0201F and BO0213D) are plotted with negative abundance. Molecular compositions in both samples with the same mass and similar normalized relative abundance are reduced toward zero. The peaks in the mass spectra are color scaled to the number of oxygen atoms in the assigned molecular formula, where it can be observed that the aged samples shift towards species with higher oxygen numbers at lower masses, compared to the fresh samples.